ships from meeting online to the first face-to-face date. It is inevitable that ... These findings align with social psychological theories of romantic attraction, which ...
A Second Chance to Make a First Impression: Factors Affecting the Longevity of Online Dating Relationships Lindsay Shaw Taylor1, Andrew T. Fiore2, G. A. Mendelsohn1, and Coye Cheshire2 1
Department of Psychology and 2School of Information University of California, Berkeley Tolman Hall, Berkeley, Calif. 1 {lstaylor, jermend}@berkeley.edu, 2{atf, coye}@ischool.berkeley.edu
Abstract This research explored the transition of romantic relationships from meeting online to the first face-to-face date. It is inevitable that impressions of a partner will change to some degree, but how much, and with what consequences? One hundred and fifty users of a popular online dating site participated in the study. They recalled a person whom they had met through the site, reporting their impressions of their partners from both before and after the first face-to-face meeting. We expected, based on prior research demonstrating the importance of physical attractiveness in romantic attraction both on- and offline, that changes in beliefs about partners’ physical appeal would be the most powerful predictor of relationship longevity. However, they were unrelated to relationship success. Across all the dimensions we examined, impressions were in fact relatively stable, but when respondents said they knew their partners better after meeting face-to-face, relationships lasted longer.
Introduction Unlike many forms of traditional dating that begin with face-to-face encounters, online dating relationships begin by using more limited communication channels and later transition to face-to-face. Partners form initial impressions of each other from online dating profiles, followed by chatting, email, and/or telephone, before finally arranging a face-to-face rendezvous. Therefore, partners who meet online often have well-formed impressions of one another before their first “real date,” which might or might not be corroborated upon meeting face-to-face. The goal of the current research was to investigate how impressions of partners change during that transition from online to offline dating and, specifically, how those changes affect relationship longevity. The nature of online dating is that individuals who post profiles try to strike a balance between presenting themselves truthfully and putting their best face forward. Other users who view their profiles try to form impressions of them while read-
ing between the lines (Ellison, Heino, and Gibbs 2006). Although there is some evidence that individuals can judge others’ personalities fairly well from online selfpresentations (Vazire and Gosling 2004), there is still much room for perceiver error. We know, too, that people sometimes misrepresent themselves in their dating profiles (Hancock, Toma, and Ellison 2007). Undoubtedly, then, some people are disappointed upon meeting their dates face-to-face, while others are pleasantly surprised. Research has yet to address how relationships are affected when online daters meet face-to-face, and their expectations about their partners are or are not confirmed. We were interested particularly in how changing impressions on specific dimensions (e.g., perceptions of personality) would correlate with relationship longevity. Judgments of physical attractiveness were obviously a prime candidate for study. Pictures are usually among the first pieces of information online daters receive about one another. In previous work, we have found that physical attractiveness plays a major role in attraction to and selection of romantic partners in online dating contexts. For example, the strongest predictor of whether an online dating profile will be seen as attractive overall is whether the photo is attractive; having attractive text-based components (i.e., self-descriptive content) is considerably less important (Fiore et al. 2008). Also, online daters who see themselves as attractive (Fiore and Donath 2005) and who are judged by others to be physically attractive (Lee et al 2008; Shaw Taylor et al. in prep) are contacted by more people and receive more messages overall than less attractive individuals. These findings align with social psychological theories of romantic attraction, which have long noted the central role physical attractiveness plays in initial attraction (e.g., Walster 1970; Walster et al. 1966). At the same time, if a romantic relationship is to succeed in the long term, it is clear that factors other than physical attractiveness must gain importance. Thus, in the current research, we also examined online daters’ impressions of their partners’
personalities, along with other more global judgments of their partners, as they transitioned from online to offline relationships.
Overview of the Current Research Users of a popular online dating site recalled partners whom they had met first through the site and subsequently for a face-to-face date. Because of the primacy placed on physical attractiveness in prior research, we focused first on the effect of being pleasantly or unpleasantly surprised by a date’s appearance. We hypothesized that seeing one’s date as more attractive before meeting face-toface than after would be associated with shorter relationships, while having a more positive impression of a date’s physical attractiveness after meeting face-to-face than before would predict longer relationships. We also asked participants about their impressions of their dates’ personalities, how close they were to their ideal for a mate, how well they got to know them, and their overall attraction to their dates before and after meeting face-to-face. As with physical attractiveness, we predicted that those who had more positive impressions of their partners after meeting face-to-face would have longer relationships, while those whose impressions declined would have shorter relationships. Finally, we predicted that changing impressions of a date’s physical attractiveness would be a more powerful predictor of relationship longevity than change on the other dimensions.
Method Participants Participants were active users of a major U.S. online dating site who were recruited in 2008 and 2009 via a pop-up advertisement offering the chance to participate in a psychological study of online dating. Data were collected from 150 respondents (78 female, 68 male, 4 unreported) who had gone on a face-to-face date with someone they met through the site. The average age of our participants was 44.8 years (SD = 11.1), and this sample was largely European American (n = 123). On the whole, participants were experienced online daters, with 18% reporting having used online dating for less than six months, 5.3% using online dating for six to twelve months, 12.7% for one to two years, 63.3% for more than two years.
Procedure When respondents clicked on the pop-up advertisement to get more information about the study, they were taken to an informed consent page, which provided study details and asked them to join. After consenting, participants completed a series of self-descriptions and psychometric measures. Next, we asked whether they were currently talking to someone whom they were considering meeting face-to-face. Those who said no were redirected to the current questionnaire, which asked them to rate someone
whom they had previously met through the site and with whom they had subsequently gone on at least one face-toface date. Participants provided the dates’ first names, which were inserted into the items (see blanks below). The questionnaire consisted of two sections. In the first, participants were asked to think back to when they were talking to their dates online, before they had actually met face-to-face. Participants rated their impressions of their dates at that point in the relationship using the items: On the whole, how attracted were you to ___; How much did you think you had in common with ___; How close did you think ___ was to your ideal for a partner; and How well had you gotten to know ___? Participants also rated how much their partners’ personality and physical appearance had appealed to them. Next, participants were asked to think about how they felt about their dates after their first face-to-face meeting. They then completed the same items, worded to refer to this later time point. All ratings were made using 7-point Likert-type scales. Finally, participants indicated on how many additional dates they had gone with their partners using a 5-point scale (0=no dates after the first, 1=one to two, 2=three to four, 3=five to ten, 4=more than 10) and whether they were still dating at the time of the study.
Results Because we were interested in how changing impressions are related to relationship longevity, for each of the six pre-date and post-date items, we computed change scores by subtracting pre-date ratings from post-date ratings. In each case, positive values indicate a more positive impression of one’s partner after the face-to-face meeting, while negative scores indicate more negative impressions after the face-to-face meeting. Examining the descriptive statistics (see Table 1) it is clear that both pre-date and post-date impressions were positive (that is, above the scale midpoint of 3 on the 0–6 scales). We did not see particularly large declines in participants’ impressions of their partners over the online- tooffline transition, despite prior evidence that people’s impressions of online partners might be more optimistic Pre-date rating
Post-date rating
Diff. score
Overall attraction
4.22 (1.17)
4.18 (1.79)
-.05 (1.61)
Have in common
4.06 (1.07)
3.92 (1.54)
-.13 (1.48)
Close to ideal
3.53 (1.33)
3.41 (1.82)
-.12 (1.52)
Know well
3.42 (1.35)
3.96 (1.85)
.54 (2.02)
Personality appeal
4.49 (1.04)
4.46 (1.52)
-.05 (1.32)
Appearance appeal 4.35 (1.30)
4.21 (1.75)
-.10 (1.42)
Table 1. Means and standard deviations (in parentheses). Ratings were made using 0-6 scales.
than they would be in a richer medium (Walther 1996; Walther et al 2001). While participants’ impressions of their partners declined slightly on five of the dimensions, repeated measures ANOVAs showed that those changes were not statistically significant, ps from .29 to .72. The only significant pre-to-post-date change was that, perhaps not surprisingly, participants felt that they had gotten to know their dates better after meeting face-to-face than before, F (1, 146) = 10.43, p < .01. To examine our central question of how changing impressions are related to relationship longevity, we began by correlating each of the six difference scores with participants’ reports of how many subsequent dates they had with their partners after the first meeting (see the first column of Table 2). Change on each of the dimensions was associated with relationship length such that having more positive impressions of one’s partner after the faceto-face meeting than before was related to going on more dates with him or her. Notably, though, the association was weakest with judgments of physical attractiveness (appearance appeal). To directly compare the relative importance of each of the dimensions, we entered the six change scores into a multiple regression model with number of subsequent dates (scaled) as the outcome variable. The results are presented in the second column of Table 2. Only change in how well participants felt they got to know their partners independently predicted relationship longevity. None of the other predictors, including physical appearance, was significant. To directly compare the predictive power of physical attractiveness and personality, we conducted a follow-up multiple regression with change on those two dimensions as the only predictors of relationship longevity. Changing judgments of personality significantly predicted number of subsequent dates, B = .30, t(133) = 3.05, p < .01; participants who had increasingly positive views of their partners’ personalities had longer relationships. Changing judgments of appearance were again unrelated to longevity, B = .08, t < 1, ns. Post-date minus pre-date change scores
Correlation with number of subsequent dates
Standardized beta
Overall attraction
.37 ***
-.027
Have in common
.44 ***
.152
Close to ideal
.50 ***
.211
Know well
.53 ***
.294 **
Personality appeal
.36 ***
.065
Appearance appeal
.26 **
-.059
Table 2. Association between changing impressions and relationship longevity. Standardized betas are from a multiple regression model predicting number of subsequent dates with all six change scores entered simultaneously. *** p < .001, ** p < .01, * p < .05
Post- minus predate change scores
Mean (SD), ongoing relationships
Mean (SD), terminated relationships
t
Overall attraction
.53 (.80)
-.14 (1.70)
2.73 *
Have in common
.65 (1.22)
-.27 (1.50)
2.40 *
Close to ideal
.71 (1.26)
-.27 (1.53)
2.53 *
Know well
1.06 (.97)
.39 (2.06)
2.24 *
Personality appeal
.29 (.47)
-.12 (1.39)
2.46 *
Appearance appeal
.47 (1.06)
-.18 (1.46)
1.66
Table 3. Independent samples t-tests comparing changing impressions among participants who were or were not still dating their partners. * p < .05
Finally, we compared mean change scores on each of the six dimensions for participants who reported that they were still in a relationship with their partners (n =17) and those who were not (n = 124). Results are presented in Table 3. For all dimensions except physical attractiveness, the mean change scores for participants in continuing relationships were significantly more positive than those for participants whose relationships had terminated. Moreover, examining the means reveals that, once again, the greatest change occurred with regard to how well participants thought they had gotten to know their partners before and after their face-to-face meeting.
Discussion The purpose of this research was to explore how online daters’ impressions of their partners change during the period when their relationships move from online to offline, and to see how change is related to whether the relationships continue and for how long. Overall, we found that participants’ views of their partners were relatively stable, indicating that the impressions individuals form online can pass the face-to-face test. In addition, two findings emerged that are particularly noteworthy. First, although much of the research on romantic attraction both online and offline focuses on the role of physical attractiveness, we found no evidence that changes in ratings of partners’ physical attractiveness are importantly related to relationship success at this stage. Our data suggest that impressions of physical attractiveness formed online are largely confirmed when partners meet face-toface. This is contrary to the popular belief that people embellish the truth when it comes to their physical attractiveness online. One possibility is that online daters have come to expect others to slightly misrepresent how attractive they are in their profiles (Hancock et al. 2007), so they overlook minor shadings that come to light upon meeting face-to-face (Fiore and Donath 2004); such discrepancies, it appears, do not matter for the relationship.
We also think it is likely that by the time people meet face-to-face, they have moved beyond the stage in which physical attractiveness is very important for determining whether the relationship continues. Second, we found that the most important factor affecting relationship longevity was how well participants felt they had gotten to know their partners. Across all our analyses, it stood out as the dimension on which participants reported the greatest change from pre- to post-date, and it was the only one consistently related to relationship length. This is worthy of empirical follow-up. For example, it would be useful to separately examine the effects of learning more factual information about one’s partner and developing intimacy with him or her. We also wonder if there is a ceiling on this effect such that too sudden an increase in familiarity might be off-putting. There is an optimal pace at which intimacy develops (e.g., Collins and Read 1990), and too great a change might have the effect of undermining the relationship. Retrospective bias is a possible limitation of this design. However, we compared the current findings to data collected from a separate sample who actually rated their dates before and after meeting face-to-face. Importantly, change scores were in the same direction and of about the same magnitude in the retrospective and contemporaneous samples on the key variables of appeal of physical attractiveness and how well participants got to know their partners. Change scores were smaller for the dimensions of overall attraction and “how much in common” when measured retrospectively, but all changes were in the same direction. Unfortunately, relationship longevity data is not available from the contemporaneous sample, so we could not test our hypotheses in that dataset. Finally, we acknowledge that although we talked about relationship longevity, these relationships were relatively short-term. However, they represent a stage of romantic relationships rarely studied, the intermediate phase between initial attraction and serious commitment. This phase has received relatively little research attention, probably due in large part to the difficulty of recruiting couples who have just started to date. Online dating provides a rare opportunity to explore how romantic relationships unfold in their early stages, but we caution against drawing generalizations to traditional dating until more is known about how they compare.
Acknowledgements This work was supported in part by the National Science Foundation, HSD-IIS 0624356.
References Collins, N. L. and Miller, L. C. 1994. Self-Disclosure and Liking: A Meta-Analytic Review. Psychological Bulletin 116: 457-475.
Ellison, N.; Heino, R.; and Gibbs, J. 2006. Managing Impressions Online: Self-Presentation Processes in the Online Dating Environment. In Journal of ComputerMediated Communication 11(2). Fiore, A.T. and Donath, J. 2004. Online Personals: An Overview. Short paper, ACM Computer-Human Interaction 2004. Fiore, A.T. and Donath, J. 2005. Homophily in Online Dating: When Do You Like Someone Like Yourself? Short paper, ACM Computer-Human Interaction 2005. Fiore, A.T.; Shaw Taylor, L.; Mendelsohn, G.A.; and Hearst, M. A. 2008. Assessing Attractiveness in Online Dating Profiles. In Proc. ACM Computer-Human Interaction 2008. Hancock, J.T.; Toma, C.; and N. Ellison. 2007. The Truth about Lying in Online Dating Profiles. In Proc.ComputerHuman Interaction 2007. Lee, L.; Loewenstein, G.; Ariely, D.; Hong, J.; and Young, J. 2008. If I'm Not Hot, are You Hot or Not? Physical Attractiveness Evaluations and Dating Preferences as a Function of Own Attractiveness. Psychological Science 19: 669-677. Shaw Taylor, L.; Fiore, A. T.; Mendelsohn, G. A.; and Cheshire, C. (in preparation). An Empirical Test of the Matching Hypo-thesis in Online Dating. Vazire, S. and Gosling, S. 2004. e-Perceptions: Personality Impressions Based on Personal Websites. Journal of Personality and Social Psychology 87: 123-132. Walster, E. 1970. The Effect of Self-Esteem on Liking for Date of Various Social Desirabilities. Journal of Experimental Social Psychology 6: 248-253. Walster, E.; Aronson, V.; Abrahams, D.; and Rottman, L. 1966. Importance of Physical Attractiveness in Dating Behavior. Journal of Personality and Social Psychology 4: 508-516. Walther, J.B. 1996. Computer-Mediated Communication: Impersonal, Interpersonal, and Hyperpersonal Interaction. In Communication Research 23: pp. 3–44. Walther, J.B.; Slovacek, C.L.; and Tidwell, L. C. 2001. Is a Picture Worth a Thousand Words? Photographic Images in Long-Term and Short-Term Computer-Mediated Communication. Communication Research 28: 105-134.
